*Article* **Vegetative Reproduction Is More Advantageous Than Sexual Reproduction in a Canopy-Forming Clonal Macroalga under Ocean Warming Accompanied by Oligotrophication and Intensive Herbivory**

**Hikaru Endo 1,2,\* , Toru Sugie <sup>1</sup> , Yukiko Yonemori <sup>1</sup> , Yuki Nishikido <sup>1</sup> , Hikari Moriyama <sup>1</sup> , Ryusei Ito <sup>3</sup> and Suguru Okunishi <sup>1</sup>**


**Citation:** Endo, H.; Sugie, T.; Yonemori, Y.; Nishikido, Y.; Moriyama, H.; Ito, R.; Okunishi, S. Vegetative Reproduction Is More Advantageous Than Sexual Reproduction in a Canopy-Forming Clonal Macroalga under Ocean Warming Accompanied by Oligotrophication and Intensive Herbivory. *Plants* **2021**, *10*, 1522. https://doi.org/10.3390/ plants10081522

Academic Editor: Koji Mikami

Received: 30 June 2021 Accepted: 21 July 2021 Published: 26 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Abstract:** Ocean warming and the associated changes in fish herbivory have caused polarward distributional shifts in the majority of canopy-forming macroalgae that are dominant in temperate Japan, but have little effect on the alga *Sargassum fusiforme*. The regeneration ability of new shoots from holdfasts in this species may be advantageous in highly grazed environments. However, little is known about the factors regulating this in *Sargassum* species. Moreover, holdfast tolerance to high-temperature and nutrient-poor conditions during summer has rarely been evaluated. In the present study, *S. fusiforme* holdfast responses to the combined effects of temperature and nutrient availability were compared to those of sexually reproduced propagules. The combined effects of holdfast fragmentation and irradiance on regeneration were also evaluated. Propagule growth rate values changed from positive to negative under the combination of elevated temperature (20 ◦C–30 ◦C) and reduced nutrient availability, whereas holdfasts exhibited a positive growth rate even at 32 ◦C in nutrient-poor conditions. The regeneration rate increased with holdfast fragmentation (1 mm segments), but was unaffected by decreased irradiance. These results suggest that *S. fusiforme* holdfasts have a higher tolerance to high-temperature and nutrient-poor conditions during summer than propagules, and regenerate new shoots even if 1-mm segments remain in shaded refuges for fish herbivory avoidance.

**Keywords:** climate change; foundation species; fucoid brown algae; non-additive effect; simulated herbivory

#### **1. Introduction**

Plant reproduction can be divided into sexual and asexual reproduction; the latter includes somatic embryogenesis and vegetative reproduction, which occurs without embryo formation [1]. Vegetative reproduction is common in clonal plants, which produce new shoots (i.e., ramets) from roots, stolons, or rhizomes [2–4]. The newly produced ramets obtain resources, such as nutrients and carbohydrates, from physiologically integrated mother plants through connections [4]. Moreover, fragmentation of these ramets via the disturbance or senescence of these connections often enhances new ramet production [2,3]. Therefore, vegetative reproduction may be more advantageous than sexual reproduction under resource-limited and highly disturbed environments.

Brown, red, and green macroalgae are the major primary producers in coastal marine ecosystems. Specifically, the canopy-forming large brown algae, kelp (Laminariales) and fucoid species (Fucales), are highly productive and act as foundation species [5], providing food, habitats, and spawning grounds for various marine organisms [6,7]. Furthermore, the conservation and restoration of marine macroalgal forests, which export carbon to the deep sea, may contribute to the mitigation of climate change caused by an increase in the atmospheric CO<sup>2</sup> level [8,9]. However, these algal forests have been declining due to ocean warming [10]. Above-average temperatures, combined with nutrient-poor conditions during the summer, have been known to cause physiological stress in macroalgal species [11–14].

Moreover, ocean warming has caused a range expansion of tropical herbivorous fishes into temperate waters, resulting in an increase in their grazing activity, especially in ocean warming hotspots, such as the Mediterranean and southern Japan [15]. Additionally, in the western North Pacific around southern Japan, nutrient concentrations in the surface mixed layer have been declining because the mixing of nutrient-poor surface water and nutrientrich deep water has been suppressed by ocean warming or longer-term natural climate change [16]. Consequently, the majority of the kelp and fucoid species that are dominant in temperate Japan have shifted their distributional range towards the pole [17]. However, such poleward range shifts have not been observed in the fucoid *Sargassum fusiforme* [17], implying that this species might have reproductive traits that allow for its survival in warm, nutrient-poor, and highly grazed environments in southern Japan.

*Sargassum* species generally have perennial holdfasts (analogous to rhizoids), stipes (analogous to stems), and annual shoots (i.e., main branches), which show large seasonal variations in biomass and length, with the exception of annual species such as *S. horneri* [18–21]. In temperate *Sargassum* species, including *S. fusiforme*, these shoots commonly germinate from stipes during summer, grow between autumn and spring, and decay during the subsequent summer after the production of propagules via sexual reproduction [18,20]. Moreover, vegetative reproduction via the regeneration of new shoots from holdfasts has been reported in several *Sargassum* species, including *S. fusiforme* [22–24]. In southern Japan, where fish herbivory is intensive between summer and autumn, and is weaker during winter [25], *Sargassum* shoots derived from holdfasts or propagules only grow from winter to spring, and decay during summer [26–28]. Therefore, these propagules and holdfasts, rather than the shoots, are exposed to warm and nutrient-poor conditions during summer. Previous studies have shown the effect of increased temperature on propagule growth [29,30], and the combined effects of temperature and nutrient availability or salinity on shoot growth [31–33] in *Sargassum* species. However, the combined effects of elevated temperature and reduced nutrient availability on the growth of propagules and holdfasts have rarely been evaluated; therefore, it is unclear whether sexual or vegetative reproduction is more advantageous under warm and nutrient-poor environments.

Moreover, Ito et al. [22] showed that the regeneration of new shoots from holdfasts was enhanced by cutting the filamentous holdfasts into segments <2.5 mm in length in *S. fusiforme*. This implies that this species might regenerate even after the holdfasts are fragmented by fish herbivory, as reported in *S. swartzii* [24]. They also reported that the percentage of *S. fusiforme* holdfasts that regenerated new shoots tended to increase in response to elevations in temperature (from 17 ◦C to 23 ◦C) and irradiance (from 50 to 230 µmol photons m−<sup>2</sup> s −1 ) [22]. However, the effects of a broader range of temperatures and nutrient availability on regeneration have not been studied and therefore the most important factor regulating regeneration is unclear.

Furthermore, microtopographic refuges, such as crevices, are known to enhance the recruitment and survival of *Sargassum* propagules in tropical regions, where intensive fish herbivory occurs [34]. Although these microhabitats are predicted to act as holdfast refuges from fish herbivory, reduced light availability in a shaded crevice may antagonize the positive effect of holdfast fragmentation by fish herbivory. However, the combined effects of fragmentation and decreased irradiance on regeneration are unclear based on the results of single-factor studies.

*Sargassum fusiforme* is common between lower intertidal and upper subtidal reefs in Japan, China, and Korea [35,36]. This species is edible and has been cultivated in these countries. Due to its commercial importance, several studies on the ecological and physiological traits of this species have been conducted [20,22,33,37], although the effects of ocean warming, nutrients, and herbivory on its reproductive traits have not been examined. Novel knowledge of the reproductive traits of this species may improve seeding methods for its cultivation.

In the present study, four laboratory culture experiments of *S. fusiforme* were conducted to examine (1) the combined effects of temperature (10 ◦C–30 ◦C) and nutrient availability on propagule growth, (2) the combined effects of temperature (15 ◦C–30 ◦C) and nutrient availability on the growth and regeneration rates of holdfasts, (3) the effect of high temperature (30 ◦C–38 ◦C) on the growth and regeneration rates of holdfasts, and (4) the combined effects of holdfast fragmentation and irradiance on the growth and regeneration rates of holdfasts in this species.

#### **2. Results**

#### *2.1. Experiment 1: Combined Effects of Temperature and Nutrients on Propagules*

Mean (± standard deviation) dissolved inorganic nitrogen (DIN) concentrations in 10% Provasoli's enriched seawater (PESI), 5% PESI, and sterile natural seawater (SSW) were 106.36 ± 1.07 µM, 74.89 ± 7.63 µM, and 6.94 ± 0.17 µM, respectively. The two-way analysis of variance (ANOVA) detected significant effects of temperature and nutrients, and their interaction on the relative growth rate of propagules (Table 1). The results of Tukey's test showed that there was no significant difference in the growth rate among temperatures in 10% PESI treatments, whereas the values decreased in response to elevated temperature from 20 ◦C to 30 ◦C in the 5% PESI treatment (Figure 1). The growth rate was significantly lower in non-enriched SSW treatments than in 10% PESI treatments at all temperatures. Moreover, the positive growth rate became negative in response to temperature elevation from 20 ◦C to 30 ◦C in the SSW treatment. – – – 10% Provasoli's enriched sea

**Table 1.** Results of two-way ANOVA on the effects of temperature and nutrient availability on the relative growth rate of *Sargassum fusiforme* germlings.


**Figure 1.** Relative growth rate of *Sargassum fusiforme* propagules cultured in nine different treatments (Mean + SD, *n* = 6). Different small letters indicate statistical significance among different treatments (*p* < 0.05).

#### *2.2. Experiment 2: Combined Effects of Temperature and Nutrients on Holdfasts*

The mean (± standard deviation) DIN concentrations in 25% PESI and SSW were 150.47 ± 1.18 µM and 5.26 ± 2.70 µM, respectively. The holdfast relative growth rate was significantly affected by temperature and nutrient availability, although their interaction was not significant (Table 2). The growth rate was significantly lower at 30 ◦C than at 20 ◦C, and lower in SSW treatments than in 25% PESI treatments (Figure 2). In contrast, the relative regeneration rate was unaffected by both temperature and nutrients, although values tended to decrease in response to elevated temperature from 20 ◦C to 30 ◦C, especially in the 25% PESI treatment.

**Table 2.** Results of two-way ANOVA on the effects of temperature and nutrient availability on the relative growth and regeneration rates of *Sargassum fusiforme* holdfasts.


\* Statistical significance.

**Figure 2.** Relative growth and regeneration rates of *Sargassum fusiforme* holdfasts cultured in eight different treatments (Mean + SD, *n* = 6). Different small letters indicate statistical significance among different temperature treatments (*p* < 0.05).

#### *2.3. Experiment 3: Effect of High Temperature on Holdfasts*

A significant effect of temperature (between 30 ◦C and 38 ◦C) on the relative growth rate of holdfasts was detected by one-way ANOVA (df = 4, MS = 0.859, *F* = 2.998, *p* = 0.040). Tukey's test indicated that the growth rate was higher at 30 ◦C and 32 ◦C than at 36 ◦C

under a significance level of *p* < 0.1 (Figure 3), although the differences among temperatures were not detected under a *p* < 0.05 level. The mean value was positive at 30 ◦C–32 ◦C and was negative at 34 ◦C–38 ◦C. In contrast, the relative regeneration rate was not significantly affected by temperature (df = 4, MS = 0.306, *F* = 0.75, *p* = 0.567). Regeneration was even observed at 30 ◦C and 34 ◦C, but not in any other treatments. – –

0.040). Tukey's

#### *2.4. Experiment 4: Combined Effects of Fragmentation and Irradiance on Holdfasts*

−2 −1 The holdfast relative growth rate was significantly affected by irradiance, but not by fragmentation or their interaction (Table 3). Values decreased in response to reduced irradiance from 130 to 30 µmol photons m−<sup>2</sup> s −1 (Figure 4). In contrast, the relative regeneration rate was significantly affected by fragmentation, but not by irradiance or their interaction (Table 3). Fragmentation significantly increased the regeneration rate, even in the low-irradiance treatments (Figure 4).

**Table 3.** Results of two-way ANOVA on the effects of fragmentation and irradiance on the relative growth and regeneration rates of *Sargassum fusiforme* holdfasts.


\* Statistical significance.

**Figure 4.** Relative growth and regeneration rates of *Sargassum fusiforme* holdfasts cultured in eight different treatments (Mean + SD, *n* = 6) at 30 ◦C using SSW as the culture media.

#### **3. Discussion**

−2 −1 – – − −1 Baba [29] examined the combined effects of seven temperature levels (10 ◦C, 15 ◦C, 20 ◦C, 25 ◦C, 30 ◦C, 32 ◦C, and 34 ◦C) and four irradiance levels (10, 25, 100, and 180 µmol photons m−<sup>2</sup> s −1 ) on the relative growth rate of *S. fusiforme* propagules in a 20-d experiment using 100% PESI (DIN = ca. 800 µM) as a culture medium. Growth rates were reported to be the highest at 25 ◦C–30 ◦C at 100–180 µmol photons m−<sup>2</sup> s −1 . The present 21-d study evaluated the combined effects of three temperature levels (10 ◦C, 20 ◦C, and 30 ◦C) and three nutrient levels (10% PESI, 5% PESI, and SSW) on the growth rate of *S. fusiforme* propagules, and found a significant interaction between temperature and nutrient availability. Negative effects of increased temperature from 20 ◦C to 30 ◦C were not detected in 10% PESI treatments (DIN = ca. 100 µM), but were found in 5% PESI treatments (DIN = ca. 75 µM). Moreover, the growth rate in non-enriched SSW treatments (DIN = ca. 7 µM) was lower than those in 10% PESI treatments at all temperatures, and changed from positive to negative values in response to a temperature elevation from 20 ◦C to 30 ◦C. These results suggest that the high-temperature tolerance of *S. fusiforme* propagules strongly depends on nutrient availability. Similar results have been obtained in our previous studies using juvenile sporophytes of the kelps *Undaria pinnatifida*, *Ecklonia cava*, and *Sacchrina japonica* [12–14]. Therefore, the early life stages of kelp and fucoid species appear to be vulnerable to reduced nutrient availability, especially under warm conditions, probably because of their small size and limited resource accumulation.

Yatsuya et al. [23] reported that holdfasts of *Sargassum piluriferum* and *S. alternatopinnatum* regenerated new shoots after incubation at 32.5 ◦C for 5–17 d. In the present study, there were significant effects of temperature and nutrient availability on the growth of *S. fusiforme* holdfasts, although there was no significant interaction between the two factors, indicating an additive effect. Holdfast growth decreased in response to elevated temperature from 20 ◦C to 30 ◦C and reduced nutrient availability from 25% PESI (DIN = ca.150 µM) to SSW (DIN = ca.5 µM). However, the growth rate remained at positive values even at 30 ◦C in nutrient-poor SSW treatment, in contrast to the results of the propagules. More-

over, the growth rates maintained positive values at 32 ◦C, and became negative at 34 ◦C in our subsequent 28-d experiment using SSW as a culture medium, whereas the propagules decreased their growth rate in response to the temperature elevation from 30 ◦C to 32 ◦C and withered within 4 d at 34 ◦C under nutrient-rich 100% PESI conditions [29]. These results suggest that the holdfasts of this species have a higher tolerance to high-temperature and nutrient-poor conditions during summer than propagules of the same species. Hightemperature tolerance of marine macroalgae is known to be associated with the ability to accumulate and maintain an internal nitrogen reserve [11,38,39]. Hence, the holdfasts, which are larger than the propagules, may withstand warm and nutrient-poor conditions using stored nitrogen.

Ito et al. [22] showed that the number of regenerated shoots per holdfast length was higher for holdfasts that were fragmented into lengths of 1 mm than for the longer holdfasts (i.e., 2.5, 5, 10, and 20 mm in length) in *S. fusiforme*. Similarly, in the present study, the regeneration rate significantly increased with holdfast fragmentation from 5 mm to 1 mm in length. Ito et al. [22] also reported that the percentage of *S. fusiforme* holdfasts that geminated new shoots tended to decline in response to decreased temperature (from 23 ◦C to 17 ◦C) and irradiance (from 230 to 50 µmol photons m−<sup>2</sup> s −1 ). However, in the present study, the regeneration rates of new shoots from *S. fusiforme* holdfasts were not significantly affected by broader ranges of temperature (15 ◦C–30 ◦C and 30 ◦C–38 ◦C), nutrient availability, or irradiance, although the values tended to decrease in response to temperature elevation from 20 ◦C and 30 ◦C and nutrient enrichment. These results indicated that the regeneration of new shoots from holdfasts of this species is strongly regulated by physical stimulation (i.e., fragmentation) rather than abiotic environmental factors. This enhancement of vegetative reproduction by fragmentation is common in clonal plants and may be advantageous for reproduction in highly disturbed environments [2,3].

Loffler et al. [24] showed that the survival rate of *S. swartzii* was unaffected by the experimental removal (to mimic fish herbivory) of 50% of holdfast biomass but decreased when 75% of holdfast biomass was removed. In contrast, Ito et al. [22] and the present study showed that the filamentous holdfasts of *S. fusiforme* regenerated new shoots even after they were fragmented into 1-mm segments. Moreover, the present study showed that the regeneration rate was unaffected by reduced irradiance from 130 to 30 µmol photons m−<sup>2</sup> s −1 , and the positive effect of fragmentation was not antagonized by the irradiance reduction. Hence, the holdfasts of this species can regenerate new shoots if the tiny segments remain in shaded refuges, such as crevices [34], to avoid intensive fish herbivory. However, holdfast growth decreased in response to reduced irradiance in the present study. Therefore, the survival rate of the holdfasts may depend on the light environment of these refuges.

*Sargassum fusiforme* is distributed from Hokkaido in northern Japan to Okinawa Prefecture in southern Japan [35]. The maximum seawater temperature and DIN concentration ranges during summer (between July and September) are 28.2 ◦C–29.1 ◦C and 1.1–4.5 µM, respectively, at several sites in Kagoshima Prefecture [26,27], near the southern distributional limit of this species. The results of the present study predict that *S. fusiforme* holdfasts can grow during the summer in Kagoshima Prefecture, and have the potential to survive under a further warming of 2 ◦C–3 ◦C [40], whereas their propagule survival during summer depends on the local nutrient environment in this region. The present study also showed that holdfast fragmentation enhanced vegetative reproduction. These traits may allow survival under the warm, nutrient-poor, and highly grazed environments in southern Japan, where local extinctions of other kelp and fucoid species have been reported [17]. However, further genetic approaches are required to quantify the contribution of vegetative and sexual reproduction to population persistence, and to determine the genotype that enables the growth of sexually-reproduced propagules into matured individuals with large holdfasts that exhibit vegetative reproduction.

*Sargassum fusiforme* is a popular food in China, Korea, and Japan, and therefore seeding methods for the cultivation of this species have been developed [22,41–43]. Ito et al. [22]

suggested that holdfasts harvested during spring can be stored by incubation at an irradiance of 1 µmol photons m−<sup>2</sup> s −1 in order to suppress regeneration, and can be utilized for the seeds by means of a tank culture for 40 d after cutting them into segments <2.5 mm in length. This method seems to be more efficient and effective than the one based on sexual reproduction in southern Japan under warming, according to the results of the present study. However, little is known about differences in growth characteristics between regenerated shoots and sexually reproduced shoots. Further studies on the ecological and physiological traits of holdfasts and regenerated shoots of this species may provide insights for the conservation and restoration of marine macroalgal forests in southern Japan under ocean warming conditions, and for the improvement of seed production methods for this species.

#### **4. Materials and Methods**

#### *4.1. Experiment 1: Combined Effects of Temperature and Nutrients on Propagules*

*Sargassum fusiforme* propagules were cultured for 21 d in nine different treatments, consisting of three temperature levels (10 ◦C, 20 ◦C, and 30 ◦C) and three nutrient levels. These temperature levels were chosen based on the study of Baba [29], which showed that the growth rate of *S. fusiforme* propagules increased in response to elevated temperature from 10 ◦C to 25 ◦C, and was similar between 25 ◦C and 30 ◦C at an irradiance of 100 and 180 µmol photons m−<sup>2</sup> s −1 in a 20-d experiment using nutrient-rich PESI [44] as the culture medium. The three nutrient levels were set as 10% PESI, 5% PESI, and SSW in order to evaluate the effect of reduced nutrient availability compared to natural levels. The DIN concentrations in the culture media were measured using an autoanalyzer with four replications.

In detail, matured shoots of *S. fusiforme*, in which many propagules were observed on the surface of female receptacles, were collected in June 2020 from a site in Yojiro (31◦33′30" N, 130◦33′47" E), Kagoshima Prefecture, southern Japan, and were transported to the laboratory in insulated cool boxes. These shoots were placed in a plastic container containing natural seawater for 7 d, and 54 propagules, which naturally dropped from the shoots to the container bottom, were collected using a Pasteur pipette. These propagules were placed in several petri dishes (9 mm in diameter) containing 30 mL of SSW, and were incubated at a temperature of 20 ◦C and an irradiance of 30 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> with a 12 h light (L):12 h dark (D) photoperiod for 4 d until the start of the experiment.

The 54 propagules were assigned to one of the nine treatments. Each propagule was placed in six holes (one propagule per hole) of nine culture plates (P24F01S, AS ONE, Osaka, Japan) containing 2.5 mL of culture medium, and was incubated for 21 d at 130 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> with a 12 h L:12 h D photoperiod. The culture medium was changed every 3 d. Photographs of these propagules were taken under a stereoscopic microscope using a digital camera before (initial value) and after culturing (final value), and the surface areas of all propagules were measured using Image J software [45]. The initial mean surface area (± standard deviation [SD]) was 0.096 ± 0.023 mm−<sup>2</sup> . Relative growth rates (% d−<sup>1</sup> ) were calculated as 100 × ln (final value/initial value)/culture d.

All statistical analyses in the present study were performed using SPSS software version 20.0 (IBM, Armonk, NY, USA). The combined effects of temperature and nutrient availability on the relative growth rates were tested using a two-way ANOVA. When significant interactions between two factors were found, Tukey's multiple comparison tests were used to examine the differences among the nine treatments.

#### *4.2. Experiment 2: Combined Effects of Temperature and Nutrients on Holdfasts*

Holdfasts of *S. fusiforme* were cultured for 28 d in eight different treatments, consisting of four temperature levels (15 ◦C, 20 ◦C, 25 ◦C, and 30 ◦C) and two nutrient levels. The four temperature levels were chosen within the temperature range at the site of collection (ca. 15 ◦C during winter and ca. 30 ◦C during summer, H. Endo unpublished data) due to the lack of available information on the optimal holdfast growth temperature. The two nutrient levels were set at 25% PESI and SSW because the holdfast growth was very slow, and was slightly affected by nutrient enrichment using 5% and 10% PESI in our preliminary experiment. The DIN concentrations in the culture media were measured in the same manner as in experiment 1.

In detail, six *S. fusiforme* individuals with relatively large holdfasts were collected in May 2018 from the site in Yojiro, and 48 holdfast segments, 5 mm in length without shoots, were cut from the specimens (eight segments per individual). Each of the eight segments derived from an individual plant were placed in a petri dish (six dishes in total) and these dishes were incubated for 24 h at 20 ◦C and 130 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> with a 12 h L:12 h D photoperiod.

At the beginning of the experiment, the wet weight of each segment (initial value) was measured using an electronic balance (0.1 mg accuracy) after the removal of excess moisture by blotting on paper towels. The initial mean wet weight (± SD) was 19.04 ± 6.49 mg. The 48 segments were randomized into eight groups of six specimens with a similar size distribution. The eight groups were each subjected to one of the eight different treatments. These specimens were placed in a petri dish (one segment per dish) containing 30 mL of culture medium and were maintained in incubators at 130 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> with a 12 h L:12 h D photoperiod for 28 d. The culture medium in each dish was changed every 7 d. At the end of the experiment, the wet weight and number of newly regenerated shoots of the cultured segment (final value) were determined. Relative growth and regeneration rates (% d−<sup>1</sup> ) were calculated as 100 × ln (final value/initial value)/culture d. The combined effects of temperature and nutrients on the relative growth and regeneration rates were tested using two-way ANOVA and Tukey's multiple comparison tests.

#### *4.3. Experiment 3: Effect of High Temperature on Holdfasts*

*Sargassum fusiforme* holdfasts were cultured at five different temperatures (30 ◦C, 32 ◦C, 34 ◦C, 36 ◦C, and 38 ◦C). Six *S. fusiforme* individuals were collected in July 2018 and 30 holdfast segments, 5 mm in length, were cut from the specimens (five segments per individual). These segments were incubated for 24 h at 20 ◦C and 130 µmol photons m−<sup>2</sup> s −1 with a 12 h L:12 h D photoperiod. The 30 segments were randomized into five groups of six specimens with a similar size distribution. The five groups were each subjected to one of the five temperatures. These specimens were placed in a petri dish (one segment per dish) containing 30 mL of SSW and were incubated at 130 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> with a 12 h L:12 h D photoperiod for 28 d. The culture medium in each dish was changed every 7 d. The wet weight and shoot number of each segment were determined before and after culturing, and the relative growth and regeneration rates were calculated in the same manner as in experiment 2. The initial mean wet weight (± SD) was 16.11 ± 3.50 mg. The effect of temperature on relative growth and regeneration rates were tested using one-way ANOVA and Tukey's multiple comparison tests.

#### *4.4. Experiment 4: Combined Effects of Fragmentation and Irradiance on Holdfasts*

Holdfasts of *S. fusiforme* were cultured for 28 d in four different treatments, consisting of two fragmentation levels (cutting into five segments of 1 mm in length and a control without cutting) and two irradiance levels (30 and 130 µmol photons m−<sup>2</sup> s −1 ). The two fragmentation levels were selected based on a study by Ito et al. [22], which reported that the number of regenerated shoots per unit length of holdfasts was higher for the holdfasts that were fragmented into 1-mm lengths, than for the longer holdfasts (i.e., 2.5, 5, 10, and 20 mm in length). The two irradiance levels were chosen because the relative growth rate of the holdfasts was significantly affected by reduced irradiance from 130 to 30 µmol photons m−<sup>2</sup> s −1 in our previous experiment.

Six *S. fusiforme* individuals were collected in May 2019, and 24 holdfast segments were cut (four segments per individual). These segments were incubated for 24 h at 20 ◦C and 130 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> with a 12 h L:12 h D photoperiod. The 24 segments were randomized into four groups of six specimens with a similar size distribution. The four

groups were each subjected to one of the four treatments. These specimens were placed in a petri dish (one segment per dish) containing 30 mL of SSW and were incubated under a 12 h L:12 h D photoperiod for 28 d. The culture medium in each dish was changed every 7 d. The wet weight and shoot number of each segment were measured before and after culturing and the relative growth and regeneration rates were calculated in the same manner as in experiments 2 and 3. The initial mean wet weight (± SD) was 11.52 ± 4.43 mg. The combined effects of fragmentation and irradiance on relative growth and regeneration rates were tested using two-way ANOVA.

**Author Contributions:** Conceptualization, H.E.; methodology, H.E., R.I. and S.O.; validation, H.E.; formal analysis, H.E. and S.O.; investigation, T.S., Y.Y., Y.N. and H.M.; data curation, H.E.; writing original draft preparation, H.E.; writing—review and editing, H.E., R.I. and S.O.; visualization, T.S., Y.Y., Y.N. and H.E.; supervision and project administration, H.E. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was partially supported by a JSPS KAKENHI Grant (Number 20K06187).

**Acknowledgments:** We thank R. Terada of Kagoshima University and T. Noda of Japan Fisheries Research and Education Agency for their helpful comments on the manuscript. We also thank M. Matsuoka of Kagoshima University and S. Ninomiya of Kagoshima City Fisheries Cooperative for their assistance with sample collection.

**Conflicts of Interest:** All authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

#### **References**


## *Article* **Different Growth and Sporulation Responses to Temperature Gradient among Obligate Apomictic Strains of** *Ulva prolifera*

**Yoichi Sato 1,2,\*, Yutaro Kinoshita 1,3, Miho Mogamiya <sup>1</sup> , Eri Inomata <sup>1</sup> , Masakazu Hoshino <sup>4</sup> and Masanori Hiraoka 3,\***


**Abstract:** The green macroalga *Ulva prolifera* has a number of variants, some of which are asexual (independent from sexual variants). Although it has been harvested for food, the yield is decreasing. To meet market demand, developing elite cultivars is required. The present study investigated the genetic stability of asexual variants, genotype (*hsp90* gene sequences) and phenotype variations across a temperature gradient (10–30 ◦C) in an apomictic population. Asexual variants were collected from six localities in Japan and were isolated as an unialgal strain. The *hsp90* gene sequences of six strains were different and each strain included multiple distinct alleles, suggesting that the strains were diploid and heterozygous. The responses of growth and sporulation versus temperature differed among strains. Differences in thermosensitivity among strains could be interpreted as the result of evolution and processes of adaptation to site-specific environmental conditions. Although carbon content did not differ among strains and cultivation temperatures, nitrogen content tended to increase at higher temperatures and there were differences among strains. A wide variety of asexual variants stably reproducing clonally would be advantageous in selecting elite cultivars for long-term cultivation. Using asexual variants as available resources for elite cultivars provides potential support for increasing the productivity of *U. prolifera*.

**Keywords:** macroalga; *Ulva prolifera*; obligate asexual strain; relative growth rate; sporulation; land-based cultivation; germling cluster method

#### **1. Introduction**

The green macroalga *Ulva prolifera* O.F. Müller, 1778 (Class Ulvophyceae) is an example of an alga showing isomorphic alternation of generations, with sporophytes and gametophytes that are morphologically indistinguishable. In the life cycle of *Ulva*, gametophytes of two mating types release biflagellate gametes with positive phototaxis, and the zygote develops into the sporophytic phase [1]. Sporophytes release quadriflagellate meiospores through meiosis, which develop into genetically separate gametophytes [2]. In addition to the sexual life cycle, several *Ulva* species, including *U. prolifera*, are known to have two types of obligate asexual life cycles without sexual reproduction via meiosis and conjugation, reproducing through biflagellate or quadriflagellate diploid zoids specialized for asexual development, these zoids have negative phototaxis [3]. These asexual zoids were termed "zoosporoids" [4,5]. The quadriflagellate zoosporoids of obligate asexual life history were a length of <10 µm; these were smaller in size than quadriflagellate meiospores of sexual life history (>11 µm length). On the other hand, biflagellate zoosporoids of obligate asexual

**Citation:** Sato, Y.; Kinoshita, Y.; Mogamiya, M.; Inomata, E.; Hoshino, M.; Hiraoka, M. Different Growth and Sporulation Responses to Temperature Gradient among Obligate Apomictic Strains of *Ulva prolifera*. *Plants* **2021**, *10*, 2256. https://doi.org/10.3390/plants 10112256

Academic Editor: Koji Mikami

Received: 29 September 2021 Accepted: 19 October 2021 Published: 22 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

life history were a length of 8–9 µm; these were distinguished from biflagellate games (6–7 µm length) [6]. These asexual variants are regarded as diploid thalli because the amount of DNA in the cells of asexual thalli is similar to that in the cells of the sporophytic thallus [6,7]. A recent study conducting genome and transcriptome analyses of *U. prolifera* suggests that the asexual thalli originally evolved via apomeiosis in sporophytic thalli [8]. Another study has revealed that the two asexual variants of *U. prolifera* show high levels of heterozygosity in the *hsp90* gene, probably as a result of hybridization among genetically distinct gametophytes [9]. These findings indicate that evolution of the asexual variants of *U. prolifera* has been independently repeated from various genetically distinct sexual populations. Therefore, although asexual *U. prolifera* variants show obligate clonal reproduction, they may include a wide range of genotypes with various physiological characteristics.

The industrial use of *U. prolifera* has a long history, being harvested and used as food in Japan since the 10th century A.D. [10]. Since the 1990s, the artificial seedling method has been developed and the cultivation in estuaries and brackish water areas has been carried out, mainly around Shikoku Island, the 4th largest island of the Japanese Archipelago [11]. However, both cultivated and natural harvest yields have declined markedly due to various environmental factors, particularly salinity and temperature (e.g., [12]). To meet market demand, the "germling cluster" (GC) method was developed as a novel way to raise seedlings and land-based cultivation on an industrial scale began in Japan early this century [13]. Globally, the land-based cultivation of *Ulva* species is focused on biomass production in view of their rapid growth and lack of any requirements for freshwater resources or particular soils for cultivation [14–16]. However, it is important to develop a stable cultivation technique in order to achieve sustainable industrial production.

Developing an elite cultivar of *Ulva* species is an essential factor for successful landbased cultivation on a commercial scale, with, in particular, a requirement for a superior growth rate [17]. Since growth rates can vary substantially between cultivars of *Ulva*, both from within the same species [18–20] and among different *Ulva* species [21], it is important to compare and select from a wide variety of species and cultivars for optimal biomass production [16]. Asexual variants are particularly appropriate for the selection of elite cultivars because their genetic characteristics can be preserved even across many generations, which is very different from the F<sup>1</sup> hybridization techniques used with terrestrial crops. Understanding the growth characteristics of asexual variants with respect to environmental differences is also necessary in order to be able to select elite cultivars appropriate to each cultivation area. Additionally, *U. prolifera* thalli often produce and release zoids apically and become shorter in length at 20 ◦C and over [22]. Therefore, cultivars continuing vegetative growth without sporulation at 20 ◦C and over are appropriate for land-based cultivation. In short, to select elite cultivars it is important to verify the growth characteristics of different variants not only to optimize growth but also with regard to temperature effects on maturation. However, there have been few studies so far concerning the phenotypic differentiation of *U. prolifera* asexual variants.

The aim of the present study, therefore, is to identify the physiological characteristics for growth and sporulation responses of asexual variants of *U. prolifera* among recognized strains. Asexual thalli were collected from six localities in Japan and compared for their growth rate, carbon and nitrogen contents, and sporulation responses across a range of temperatures using the GC method for seedling production and cultivation to apply the industrial aquaculture perspective.

#### **2. Results**

#### *2.1. Molecular Analysis*

The *hsp90* gene sequences of the six strains showed the presence of alternative bases at some positions, indicating the presence of different alleles (Figure 1) and suggesting that these strains are heterozygous and diploid.


**Figure 1.** Comparison of *hsp90* gene sequences from *Ulva prolifera* thalli (undergoing an asexual life cycle) collected from six Japanese strains (see Table 1). Dots indicate identity with Strain 1; blanks indicate deletions; and double rows of dots and the following characters indicate the presence of different alleles: K, either G or T; M, either A or C; R, either A or G; S, either C or G; W, either A or T; Y, either C or T.


**Table 1.** Basic information on the *Ulva prolifera* strains collected from six estuarine localities in Japan for use in this study.

The Oboro River is on Hokkaido Island, the Takeshima River is on Shikoku Island, and the remaining rivers are in northeastern Honshu.

#### *2.2. Growth Rate and Sporulation at Different Temperatures*

The relative growth rate (RGR) at different cultivation temperatures differed among strains. The mean RGR of Strain 1 varied narrowly over the range 0.3–0.4 and was not significantly influenced by temperature (*p* = 0.298, Figure 2, Strain 1), and the values of the maximum were about 1.3 times those of the minimum. However, those of other localities varied among temperature and there were significantly differences by post-hoc tests (*p* < 0.01, Figure 2, Strain 2–6). Although no clear peaks of RGR of Strain 2 and 4–6 were detected, maximum values were detected at 20–30 ◦C (Figure 2, Strains 2 and 4–6). The maximum values were 2.2–2.9 times those of the minimum. However, RGRs for Strain 3 indicated a clear peak with a mean of 0.55 at 20 ◦C, which is 3 times faster than the value at 10 ◦C (Figure 2). The RGRs of Strains 1 and 3 were significantly higher than those of other strains at 10 ◦C (*p* < 0.05) and 20 ◦C (*p* < 0.01), respectively (Supplemental Table S1). Throughout the cultivation period, no sporulating cells were detected in any of the thalli incubated at 10 or 15 ◦C (Figure 3). However, from the 2nd day of cultures at 20 ◦C or above, sporulating cells were already present in Strain 4 thalli (Figure 3), and in the thalli of Strains 1–3, and 5 from the 4th day of culture (Figure 3). In Strains 3–5, sporulating cells occupied more than half the total thallus area at 30 ◦C (data not shown). Strain 6 thalli showed no evidence of sporulation in cultures incubated below 25 ◦C, sporulation only began from the 8th day of culture (Figure 3) and was limited to only the tip of the thallus (data not shown).

**Figure 2.** Relative growth rate (RGR) of *Ulva prolifera* thalli from each of six localities incubated in vitro at one of five different temperatures. The RGRs (*n* = 5) were calculated from four consecutive samples linearly arranged between 0.01 and 0.1 g (Supplemental Figure S1). Error bars indicate standard error of the mean. Different lowercase letters indicate significant differences (*p* < 0.05) among different temperatures. The results of statistical analysis among strains at each cultivation temperature were shown in Supplemental Table S1.

#### *2.3. Carbon and Nitrogen Content at Different Temperatures*

The carbon content of thalli for all six strains ranged between 0.337 ± 0.005 and 0.396 ± 0.001 mg mg−<sup>1</sup> , with no obvious peak at any particular incubation temperature, although differences detected among incubation temperatures were significant for Strains 1–3 (Figure 4).

**Figure 3.** Changes in the frequency of occurrence of sporulating individuals of *Ulva prolifera* incubated in vitro at one of five different temperatures. Values are means ± standard error; *n* = 5 individuals.

**Figure 4.** Comparison of carbon content in six strains of *Ulva prolifera* incubated in vitro at one of five different temperatures. Values are means ± standard error; *n* = 5 individuals. Different lower-case letters indicate significant differences (*p* < 0.05) among different temperatures. The results of statistical analysis among strains at each cultivation temperature were shown in Supplemental Table S1.

Nitrogen content ranged from 0.035 ± 0.001 (at 10 ◦C) to 0.055 ± 0.001 mg mg−<sup>1</sup> (at 30 ◦C), tending to increase with increasing temperature, for all except Strain 6 (Figure 5). In the latter, no significant differences were detected among different temperature incubations, the values ranging between 0.037 ± 0.001 and 0.0433 ± 0.002 mg mg−<sup>1</sup> (Figure 5).

**Figure 5.** Comparison of nitrogen content in six strains of *Ulva prolifera* incubated in vitro at one of five different temperatures. Values are means ± standard error; *n* = 5 individuals. Different lowercase letters indicate significant differences (*p* < 0.05) among different temperatures. The results of statistical analysis among strains at each cultivation temperature were shown in Supplemental Table S1.

#### **3. Discussion**

The optimum temperature for growth of *U. prolifera* is known to vary according to sampling locality (strain) within the range 15–25 ◦C (reviewed by [23]), and the thalli generally mature and release zoids at 20 ◦C or higher [22]. However, previous studies did not identify the generation or type of life cycle of the specimens used in culture experiments. The present study revealed that different strains of asexual thalli of *U.* *prolifera* had differences in thermosensitivity of growth and sporulation, and that this may be connected with the presence of different *hsp90* genotypes among these different strains. The growth rate temperature optima range within 20–25 ◦C, which is similar to that reported in previous studies [23]. It is noteworthy that for most strains the growth rate showed significant thermosensitivity. However, the growth rate of Strain 1 was not significantly influenced by temperature, suggesting that it can maintain growth even at lower temperatures. Strain 1 is from the northern Pacific coast of Hokkaido, which is close to the northern limit of distribution of *U. prolifera* [24], so it might be expected to have a lower temperature tolerance than other strains. With regard to sporulation, only Strain 6 thalli did not sporulate at 20 ◦C within 12 days, demonstrating a clearly different thermosensitivity from other strains, with a greater tolerance to high temperatures for maintaining vegetative growth.

It might be expected that the differences observed in growth rates and temperature tolerance among strains are connected with the environmental characteristics of the locality from which they were obtained. However, the results of the present study indicate that strains from neighboring localities have clearly different thermosensitivity; for instance, the localities of Strains 2 and 3 are only 5 km apart. Considering the genotypic differences among strains of *hsp90*, the phenotypic differences would be caused by genetic background. This suggests that the characteristics of each strain are not necessarily closely adapted to the environmental conditions at the locality in which it is found, and it is considered that the phenotypic diversity of this alga is itself high and not dependent upon site-specific environmental conditions. Hiraoka and Higa (2016) [3] proposed that *U. prolifera* had evolved from an ancestral marine species to become a true estuarine species: firstly, the sexual population adapted to low salinity conditions, and then a number of different asexual generations arose from genetically variable sexual ancestors, with natural selection finally producing an array of specialized asexual thallus genotypes that efficiently occupy the estuarine habitat. The variable thermosensitivity of asexual variants among localities could be interpreted as the result of the evolution and adaptation processes of this alga. The variation of *hsp90* genotypes observed among the six strains may be a manifestation of the phenotypic differentiation among them. Distinguishing among the genotypes affecting the phenotypes requires further study.

Measurements of the net photosynthetic rate and RGRs of *U. prolifera* collected from green tides in China have peak at 18–22 ◦C with a marked decline at 26 ◦C [25]. However, in the present study, carbon content was not observed to vary across different culture temperatures for all strains. This demonstrates the potential for stable carbon fixation among strains of *U. prolifera* regardless of temperature fluctuations, implying also a potential for CO<sup>2</sup> mitigation by *U. prolifera* which could be calculated from yield data.

Nitrogen content, however, varied with temperature for all strains except Strain 6, despite the presence of nutrients sufficient for culture conditions. In previous studies, the nitrogen content of *U. prolifera* collected from eutrophic areas of the Yellow Sea, was reported to be 3.6% [26]. However, values of 3–4% were found in wild-collected thalli from the same locality as Strain 6, where the dissolved inorganic nitrogen concentration measured was found to be insufficient for optimum growth [27]. In the present study, nitrogen content was in the range 2.8–5.1% across different incubation temperatures, suggesting that the assimilation capacity for nitrogen is influenced by temperature. Raven and Geider (1988) previously reported that temperature influences the nutrient-uptake rates via *Q*<sup>10</sup> effects on algal metabolism [28]. The nitrogen content might therefore reflect the physiological response to differences in temperature which was a variable among the strains in the present study.

Many green algae show rapid nutrient uptake rates, contributing to the removal of excess nutrients in the water column [29,30]. According to the results of the present study, nutrient uptake kinetics might differ among strains and this may be useful for optimizing temperature-dependent quantitative removal of nitrogen from water column in land-based

cultivation. This suggests a clear future requirement to ascertain the nutrient-uptake kinetics of each strain of asexual variant.

For practicing land-based cultivation on an industrial scale, it is impractical (and uneconomic) to have to adjust the seawater temperature in the tank by external means. Therefore, seawater pumped from offshore or from saline wells, with seasonally fluctuating temperatures, needs to be used as it is for land-based cultivation. In order to improve and maximize productivity for such seasonal changes in seawater temperature, information about growth and sporulation responses of asexual variants is required, as in the present study.

Currently, several land-based cultivation facilities in southern Japan are facing decreases in productivity due to reproductive maturation and pausing of growth at 20 ◦C or higher in summer. The use of cultivars with where sporulation does not occur until much higher temperatures, such as Strain 6 in the present study, would be one means to enable stable year-round cultivation in these southern areas. In contrast, higher productivity in the low winter temperature of northern areas require strains with higher growth rates at such temperatures. From the results of the present study, the growth rates of Strains 1 and 3 were significantly higher than other strains at 10 ◦C and 20 ◦C, respectively, so these strains can be regarded as elite cultivars at those temperatures.

It might also be effective to use these cultivars seasonally according to observed changes in seawater temperature. The present study revealed that asexual variants of *U. prolifera* cover a wide phenotypic range of thermosensitivity as a result of natural selection. With regard to preserving valuable characteristics as an elite cultivar, selection from asexual variants is a useful technique, because in sexual strains the occurrence of recombination may result in the loss of the required optimal responses. Therefore, evaluating and utilizing of these asexual variants as a resource pool of candidates for elite cultivars will help to support optimization of productivity and expand the cultivable area of algae such as *U. prolifera*.

#### **4. Materials and Methods**

#### *4.1. Collection and Stock Maintenance of Thalli*

*Ulva prolifera* thalli were collected from the estuaries of six Japanese rivers (see Table 1). To confirm the life cycle of all thalli collected at each locality, sporulation and releasing zoids were conducted according to Hiraoka et al. (2003) [6]. Zoids of samples from sites 1 to 4 and 6 were found to be biflagellate. These thalli were confirmed as asexual variants, since their zoids showed negative phototaxis and were obviously bigger than both male and female gametes reported in previous studies [6]. Zoids of site 5 were quadriflagellate. Thalli cultured from the quadriflagellate site 5 zoids released the same type of quadriflagellate zoids again. More than two generations were repeated and all released quadriflagellate zoids, confirming that site 5 thalli were obligate asexual variants. A unialgal culture strain was established for each locality (Table 1, Strains 1–6) at Usa Marine Biological Institute, Kochi University. All strains were transported to the Yuriage Factory, Riken Food Co., Ltd., in Natori City, Miyagi Prefecture, and their seeding stocks for the growth and maturation experiments were prepared according to the GC method for unattached (free-floating) macroalgal culture [13]. Thallus clusters were produced according to the method of Hiraoka et al. (2020) with slight modifications [21]. Synchronous zoid formation in each strain was induced by cutting a well-developed thallus into small fragments of 1–2 mm in length, which were washed in sterilized fresh water for approximately 10 s and cultured in a Petri dish containing 40 mL Enriched Seawater (ES) medium [31] at 20 ◦C under a 12 h:12 h L:D cycle, with light of 150 µmol photons m−<sup>2</sup> s −1 . Under these conditions, thallus fragments released biflagellate (Strains 1–4 and 6) or quadriflagellate (Strain 5) zoids within 3 days.

Aliquots of zoid suspension densely concentrated using their phototactic response were placed in Petri dishes, adjusted to a density of >10<sup>4</sup> zoids per 1 mL medium, and incubated under the same conditions as above. After 3 weeks, germlings grew at a high density on the bottom of the dish and attached to one another to form aggregations with the appearance of a green mat. The aggregations were scraped off the dish without harming them, torn into numerous small clusters and cultured with gentle aeration, allowing to drift freely within the vessel. When they attained a length of 1 mm or more, they were statically stocked under weak light (<50 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> under 12 h:12 h L:D cycle) at 20 ◦C for one week until used for growth-rate and sporulation-response experiments.

#### *4.2. Molecular Analysis*

Total genomic DNA was extracted from a small fragment of living material using BT Chelex® 100 Resin (Bio-Rad Cat# 143-2832, Hercules, CA, USA). A fragment of approximately 1 cm of each sample was ground in a 2 mL tube with 100 µL of 10% Chelex solution using a homogenization pestle at room temperature and incubated at 95 ◦C for 20 min, shaken at the middle and end of the 20 min period. The mixture was then cooled and centrifuged at 4000 rpm for 2 min.

Part of the sixth exon of the *hsp90* gene sequence was amplified using the primer pair of *hsp90*-6F (5′ -GCAGACCCAGAAAGTGATCTATTAYATCA-3′ ) and *hsp90*-6R (5′ - GCAGGYTCATCCAGACTAAATCC-3′ ), as reported by Ogawa et al. (2014) [9]. PCR amplifications were carried out using KOD FX Neo (ToYoBo, Osaka, Japan) and performed using a thermal cycler for 35 cycles of denaturation at 98 ◦C for 10 s, annealing at 55 ◦C for 30 s, and extension at 68 ◦C for 30 s, followed by a final hold at 10 ◦C. PCR products were sequenced by Fasmac (Atsugi, Kanagawa, Japan).

#### *4.3. Growth Rate at Different Temperatures*

To reduce the lag phase growth of the stocked materials, hundreds of thallus clusters for each strain were pre-cultured in a round 3L-flask with continuous aeration for 7 days. The flask was filled with sterilized seawater containing half-strength ES medium [31]. Temperature and light conditions for growth of germlings were as described above. The medium was changed every day. When the thallus clusters grew to 5–10 mm in length in this pre-culture, 8–12 clusters (0.01 g-wet, Figure 6a) were transferred to 500 mL-flasks and cultured with aeration at 10, 15, 20, 25, 30 ◦C for 8 days (Figure 6b). Light was provided from an LED unit (3LH-64, NK System Co., Ltd. Osaka, Japan) at 150 µmol photons m−<sup>2</sup> s −1 with a 12 h:12 h L:D cycle in the incubator (CN-40A, Mitsubishi Electric Engineering Co., Ltd., Tokyo, Japan). Half-strength ES medium was used as culture medium and changed every day. The temperature range and light intensity were set according to the previous study about the RGR of this alga vs. abiotic conditions [21], the RGR was saturated at a light intensity of >67 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> and indicated broad ranges from 10 ◦C to 30 ◦C. The pre-culture experiment decided the enrichment medium condition; we confirmed that the half-strength of ES medium could be sufficient for the RGR by changing it daily. To determine the fresh mass of living material without causing damage by drying, thallus clusters were held between sterilized paper towels four times to carefully remove water on the surface, immediately placed in a Petri dish (9 cm in diameter) filled with half-strength ES medium on an electronic balance (0.1 mg accuracy), quantified, and returned to the same culture condition. This mass measurement was made within a few minutes at the end of the light period every day, equally spaced at 24 h intervals. Relative growth rate (RGR: the continuously accelerating growth of algae during the exponential phase; Supplemental Figure S1) was calculated using the following equation:

$$\text{RGR} = (\ln \text{W}\_1 - \ln \text{W}\_0) \text{ day}^{-1}$$

where W<sup>0</sup> is the initial fresh mass in the culture at time zero, and W<sup>1</sup> is the mass after 24 h.

**Figure 6.** Representative images of *Ulva prolifera* used for the present study. (**a**,**b**) Thalli produced by the germling cluster method for the growth experiment: (**a**) initial thalli; (**b**) thalli after 8 d of cultivation. (**c**–**e**) Thallus for sporulation experiment: (**c**) sporulating individual with sporulated cells (arrows), (**d**) dark coloration indicating formation of sporulated cells, and (**e**) zoids released from sporulating part of thallus. Solid squares in (**a**–**c**) indicate 1 cm<sup>2</sup> ; bars (**d**,**e**), 300 µm.

#### *4.4. Sporulation at Different Temperatures*

− − Between five and eight precultured thallus clusters were selected and separated individually in the center part of clusters being careful not to injure the thallus and interfere with the release the sporulation inhibitor [32]. Three intact thalli (length 1 mm) from each strain were selected and each individual placed in a separate 500 mL flask and cultured at 10, 15, 20, 25, 30 ◦C under 150 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> with a 12 h:12 h L:D cycle for 12 days, using the same incubators and LED units as for the growth rate experiments. All were incubated in half-strength ES medium renewed every 2 days. When the medium was changed, the thalli were placed in a 9 cm Petri dish filled with half-strength ES medium and the thallus surface was observed by light microscopy for the presence or absence of sporulation (Figure 6c–e) and, using a digital camera attachment, recorded as digital images.

#### *4.5. Carbon and Nitrogen Contents at Different Temperatures*

For all strains used in the growth rate experiment at different temperatures, five clusters were randomly selected after final measurements had been taken. Since the light intensity used was above the compensation irradiance for photosynthesis [23], and the culture medium (half-strength ES; approximately 420 µM as nitrate) was changed every day, the thalli were considered to be supplied with sufficient carbon and nitrogen for normal growth to occur. Seawater was carefully blotted from the thallus surface of the cluster samples, which were placed in a dry oven (EYELA WFO-500, Tokyo Rikakikai Co., Ltd., Tokyo, Japan) for 12 h at 90 ◦C. The dried thalli were then each pulverized with a pestle and mortar and carbon and nitrogen content were measured using a CHN analyzer (Flash 2000, Thermofisher Scientific, Waltham, MA, USA).

#### *4.6. Statistical Analysis*

All data are presented as mean ± S.E. Significant differences in RGR, carbon content, and nitrogen content among different cultivation temperatures and different strains were identified by the Kruskal–Wallis test followed by Steel–Dwass multiple comparison tests. A nonparametric procedure was chosen because not all of the data were normally distributed or homoscedastic.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/plants10112256/s1, Figure S1: Upper figure: Wet weight variations in thallus clusters of six strains of *Ulva prolifera* incubated in vitro at one of five different temperatures, 150 µmol photons m−<sup>2</sup> s −1 (12/12 h light/dark cycle) in half-strength ES medium with aeration for 8 d. Lower figure: the same data expressed as natural logarithms. The slopes for each combination of strain and temperature were used to generate the relative growth rate values in Figure 2. Table S1: Differences in RGR, Carbon, and nitrogen contents among five strains of *Ulva prolifera* at 10, 15, 20, 25, and 30 ◦C. Data were analyzed using Kruskal–Wallis test followed by post-hoc Scheffe's test for multiple comparisons.

**Author Contributions:** Conceptualization, Y.S.; methodology, Y.K., M.M., E.I., M.H. (Masakazu Hoshino); validation, Y.S.; formal analysis, Y.S. and E.I.; investigation, Y.S., Y.K., M.M., M.H. (Masakazu Hoshino). and M.H. (Masanori Hiraoka); data curation, Y.S. and M.H. (Masanori Hiraoka); writing—original draft preparation, Y.S.; writing—review and editing, M.H. (Masakazu Hoshino), and M.H. (Masanori Hiraoka); visualization, E.I. and Y.S.; supervision and project administration, Y.S. and M.H. (Masanori Hiraoka). All authors have read and agreed to the published version of the manuscript.

**Funding:** This work partially supported by basic-science research funding from Riken Food Co., Ltd., in 2019–2021 to Y.S., M.M., and Y.K.; and from the JST-OPERA Program (Grant Number JPMJOP1832) to Y.S., Y.K., and M.H. (Masanori Hiraoka).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data is contained within the article and supplementary material.

**Acknowledgments:** We thank H. Nagoe, M. Hoshi, and Y. Chiyokawa at the Yuriage Factory of Riken Food Co., Ltd., for their support with statistical analysis and cultivation studies. We also thank A. Kokubun and T. Taki of Riken Vitamin Co., Ltd., for their support with analysis of carbon and nitrogen contents.

**Conflicts of Interest:** All authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

#### **References**


## *Review* **Massive** *Ulva* **Green Tides Caused by Inhibition of Biomass Allocation to Sporulation**

**Masanori Hiraoka**

Usa Marine Biological Institute, Kochi University, Inoshiri, Usa, Tosa, Kochi 781-1164, Japan; mhiraoka@kochi-u.ac.jp; Tel.: +81-88-856-0462

**Abstract:** The green seaweed *Ulva* spp. constitute major primary producers in marine coastal ecosystems. Some *Ulva* populations have declined in response to ocean warming, whereas others cause massive blooms as a floating form of large thalli mostly composed of uniform somatic cells even under high temperature conditions—a phenomenon called "green tide". Such differences in population responses can be attributed to the fate of cells between alternative courses, somatic cell division (vegetative growth), and sporic cell division (spore production). In the present review, I attempt to link natural population dynamics to the findings of physiological in vitro research. Consequently, it is elucidated that the inhibition of biomass allocation to sporulation is an important key property for *Ulva* to cause a huge green tide.

**Keywords:** biomass allocation; green tide; sporulation; *Ulva ohnoi*; *Ulva prolifera*; vegetative growth

**Citation:** Hiraoka, M. Massive *Ulva* Green Tides Caused by Inhibition of Biomass Allocation to Sporulation. *Plants* **2021**, *10*, 2482. https:// doi.org/10.3390/plants10112482

Academic Editor: Koji Mikami

Received: 4 October 2021 Accepted: 13 November 2021 Published: 17 November 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

*Ulva* (Ulvophyceae, Chlorophyta) or sea lettuce is the most abundant green seaweed and is ubiquitous in tropical and temperate coastal ecosystems around the world. The genus *Ulva* currently includes at least 85 taxonomically accepted species [1]. The thallus body is uniformly sheet-like, being two cells thick or tubular with a single cell layer, except for a very small holdfast part (Figure 1A,B). The generation time of *Ulva* is short, and its spores can develop into a thallus having the potential ability to produce spores again in 2–3 weeks. Spore formation occurs in the somatic cells which directly transform into sporangia (Figure 1C,D). The sporulation first occurs at the tip of the algal body and then sequentially occurs toward the bottom, several tens of spores per sporangia are released, and the empty sporangium are spontaneously detached from the thallus.

A few *Ulva* species cause massive kilometer-scale blooms termed as green tides, and these have been recorded mainly around the industrialized coastlines of Europe, North America, and east Asia [2]. The world's most extensive *Ulva* green tides have repeatedly occurred in the Yellow Sea since 2007 [3] or 2008 [4]. The causative green tide species can substantially increase their biomass in a free-floating state by increasing the size of the thalli and their fragments. Smetacek and Zingone [5] pointed out that this is crucial because it is the unattached forms that, by invading new space, are able to increase their nutrient supply, free themselves from competition for limited hard substrates, and avoid their many benthic grazers. As a result, the unattached forms can build up a large biomass, forming massive green tides. In general, most *Ulva* spp. grow when their motile spores settle on the substrate. The attached populations in temperate regions are regularly present in early spring, increase to a maximum in late spring, and rapidly decrease through summer, showing a unimodal pattern of biomass fluctuation (Figure 2A) [6–9]. The summer decline of the attached populations occurs markedly at 20–25 ◦C. However, green tide species peak in biomass in the summer [10] and occasionally continue to grow [11]. Such a temporal difference of thallus growth pattern between attached populations and floating green tide populations has not received much attention in the literature. I consider that the species-

specific differences of ecophysiological characteristics are a crucial key in determining if an algal species causes an excessive bloom or not.

**Figure 1.** A living *Ulva* specimen (*U. aragoënsis*). (**A**) The developed thallus with sporulation in the upper part. Above the part indicated by the arrow, all the cells formed spores and some of them released spores. Arrowhead indicates a small holdfast; (**B**) Cross-section of the middle part of the thallus having a two-cell layered structure; (**C**) Surface view of somatic or blade cells in the vegetative state in the middle part of the thallus; (**D**) Surface view of the cells forming spores. Arrowheads indicate empty cells after spores were released.

**Figure 2.** Comparison of seasonal abundance change between the attached populations and the green tides of *Ulva* in the temperate region. Arrowhead indicates the period when biomass increase is suppressed mainly by light limitation. Arrows show the period when allocation of biomass to sporulation begins to be greater than vegetative growth. (**A**) The attached *U. australis* modified from [9,12] (black line). Dotted line is a growth curve predicted in case of no light limitation. *Ulva ohnoi* green tide modified from [11] (red line); (**B**) Attached *U. prolifera* in brackish water modified from [13,14] (black line). The *U. prolifera* green tide in the Yellow Sea modified from [4,15] (red line).

In various previous publications, it has been explained that *Ulva* green tides are a symptom resulting from eutrophication [16]. Indeed, the supply of dissolved inorganic nitrogen (N) and phosphorus (P) is required to sustain *Ulva* thallus growth. However, as an example of green tides in Tokyo Bay, the scale of *Ulva* blooms has expanded despite a significant decrease in N and P concentrations year by year [17]. In this case, a subtropical species, *U. ohnoi*, unintentionally introduced and excessively grew as unattached form without the summer decline. In the Yellow Sea, free-floating *Ulva* populations during the early stage of green tides in spring include four or more *Ulva* species, but only one species, *U. prolifera*, continuously expand, resulting in monospecific spectacular blooms in summer [18,19]. All the other species do not seem to be able to continue vegetative growth while enduring the high temperatures from spring to summer, even though they sympatrically experience the same eutrophic conditions in the Yellow Sea. These examples of *U. ohnoi* and *U. prolifera* indicate that when specific species, having a continuous growth ability in high temperatures, encounter the minimum nutrient conditions for sustaining its vegetative growth, a huge bloom can occur.

In this review, I compare the *Ulva* biomass fluctuation patterns between attached populations and green tide populations and explain which ecophysiological properties of a species or strain is essential for causing green tides. In addition, I will focus on the mechanism of switching between the vegetative growth and sporulation of *Ulva* cells. By relating field observations and laboratory experiments, I attempt to more comprehensively examine the relationships between the cells, individuals, and populations underlying the mechanism of the development of green tides.

#### **2. Attached Population Dynamics**

#### *2.1. Population Fluctuation Follows Individual Size Fluctuation*

In the temperate coastal zone, the biomass of attached *Ulva* populations fluctuates seasonally according to the periodic fluctuation of the water temperature. Many temperate species such as *U. rigida*, *U. lactuca*, and *U. australis* (syn. *U. pertusa*) show a regular fluctuation pattern in which the attached biomass increases from winter to spring and declines during high temperatures from summer to autumn, as described in Figure 2A [6–9,12,20,21]. The biomass peaks shift to later in the year in cold regions at high latitudes [22]. This review will progress the story about the temperate populations. There has been detailed demographic research study on the attached population of *U. australis* over a period of three consecutive years [9]. It demonstrated that the seasonal fluctuations of the *Ulva* population synchronize with those of the thallus size rather than with changes in the density of thallus individuals. This indicates that biomass fluctuations of *Ulva* population are attributed to that of well-developed thallus individuals.

#### *2.2. Increase Phase*

*Ulva* with a simple multicellular body perform 'diffuse growth' in which cell divisions can occur more or less throughout the tissues of the organism [23]. The somatic cells divide synchronously in standardized conditions once a day [24]. Therefore, the *Ulva* thalli are capable of exponential growth, displaying extremely high growth rates. In fact, a daily rate of over fourfold in *U. meridionalis* in the culture experiment has been reported as the highest growth rate ever reported for multicellular autotrophic plants. In the same paper, a strain isolated from the attached *U. prolifera* population was also revealed to display two-fold growth rate per day [25]. The high exponential growth of *Ulva* spp. generally occurs in high temperatures of 20–30 ◦C in suitable light and nutrient conditions, as described below in Section 3. If such high growth continues in the sea from spring to summer with the optimum high temperatures, a bloom would occur explosively. However, in the attached population, the rapid biomass increase is suppressed mainly by light limitation caused by self-shading as density increases (Figure 2A). Although light limitation is an inevitable suppression factor, the population is also negatively affected by some irregular factors of low salinities by precipitation and herbivory by benthic organisms such as snails and sea hares. Their inhibitory effects occur because the population is attached to the substrate.

#### *2.3. Decline Phase*

As the water temperature rises over 20 ◦C, *Ulva* thalli are highly promoted to produce and release spores. The allocation to sporulation in thalli of the attached *Ulva* populations has been observed to be significantly larger during warmer months [12,20]. Niesenbaum [20] described "the seasonality of reproduction, and changes in the abundance of total biomass and reproductive biomass during the reproductive season could have been a function of temperature. The sharp decline of total biomass in early August, and its low rate of recovery through August and September were probably due to temperature effects on growth and reproduction". Furthermore as "When temperature reached seasonal highs, allocation of biomass to the formation and release of swarmers (spores) was greatest, while the rate of vegetative replacement diminished as temperatures first became suboptimal and then inhibitory for growth. This could explain the increases in percent reproductive tissue during August and September". These findings are essential for understanding the *Ulva* biomass fluctuation. However, so far the allocation of biomass to sporulation in the decline phase of *Ulva* populations has been almost overlooked. Practically, only the intrinsic traits involved in the increase phase, such as high growth rates or multiple reproduction modes, have been highlighted [3,26]. Recently, a few works examined the decline phase of the *U. prolifera* green tide in the Yellow Sea [10]. However, little coverage has been given to the allocation to sporulation.

#### *2.4. Disappearance Phase*

After the decline phase, *Ulva* thalli almost disappear in the autumn. In this disappearance phase, although evidence has not been provided from field research yet, small individual thalli of less than a few centimeters in size could release spores and disappear, whereby their settled spores grow up fast to small thalli and release spores again in the early developmental stage. This fast generation alternation may occur until the water temperature drops below 20 ◦C in temperate species. These predictions are derived from culture work in the laboratory as described next.

#### **3. Individual Size Determined by Vegetative Growth and Sporulation Decay**

Culture experiments using temperature-controlled incubators have confirmed that higher temperatures promote sporulation decay [27]. An asexual variant of *U. prolifera* originally isolated from the attached population in brackish water and its clonal offspring thalli were tested (Figure 3). According to this study, their growth rates increase as the temperature rises to 25 ◦C. However, sporulation at the apical part of the thallus occurs earlier as the temperature rises, and the amount of sporulation decay increases. At 30 ◦C after sporulation first occurs, the thalli repeatedly produce spores and as a whole continue to decrease, resulting in the disappearance in one and half months of culture. At 20 ◦C and 25 ◦C, the vegetative growth increment and the amount of decay due to sporulation are balanced, and the total length of the thalli cannot extend from about 10 cm. At 15 ◦C, because the vegetative growth exceeds the small amount of sporulation, the thalli continue to grow larger. At the low temperature of 10 ◦C, sporulation does not occur, and the thalli continue to grow slowly. From these findings, individual thallus mass (M) can be expressed by the two factors of vegetative growth (G) and the amount allocated to sporulation (S) as M = G − S. As the *U. prolifera* strain has the intrinsic trait of S ≥ G at ≥20 ◦C after the first sporulation, M becomes constant or decreases over 20 ◦C.

**Figure 3.** Change of averaged thallus length of *Ulva prolifera* strain isolated from the attached population in the Yoshino River estuary, Japan, cultured at different temperatures. Each arrow indicates the day when sporulation first occurred. After that, sporulation occurred repeatedly. Only at 10 ◦C, no sporulation occurred. This figure was redrawn based on the data from [27].

#### **4. Inhibition of Sporulation Leads to Green Tides**

#### *4.1. Ulva ohnoi Green Tide*

≥ − Floating thalli which are free from the attached substrate can spread moderately and receive sufficient light. If the floating thalli acquires the property of inhibiting sporulation even at high temperatures, they would make a massive green tide. Here, the intrinsic trait to cause a large-scale green tide is expressed as S < G at ≥20 ◦C. Then, as S is nearly zero and G is usually the maximum growth rate that the *Ulva* species can attain, M increases exponentially. In addition to the high-temperature growth property, the structural feature of easily producing floating thalli and their fragments also promotes the magnification of the green tides. *Ulva ohnoi* is a typical species with these characteristics. When this species was reported as a new green tide-forming species, it was taxonomically described as having a large, thin, and fragile blade thallus easily torn into floating fragments in the diagnosis [28]. The field observation of the *U. ohnoi* green tide was first made in Tosa Bay, southwestern Japan, for two years [11]. It shows that the thallus fragments grew rapidly at over a five-fold growth rate per week as the summer water temperature rose to 28–30 ◦C, reaching an average length of about 50 cm, a maximum length of >1 m, and the largest biomass attained approximately 1 kg fresh mass m−<sup>2</sup> in August. In the decline phase of the green tides, a small amount of sporulation occurred in June-August, but half and more parts of the well-developed blades frequently formed and released spores in October, resulting in the population decline (Figure 2A). The remarkable difference compared to the temperate attached population is that *U. ohnoi* continues to grow with no or little biomass allocation to sporulation during the period of the summer decline. The properties are summarized below.


Of these, 1 is particularly important for the occurrence of green tides. If the species does not produce spores and does not reduce its total mass, the green tide biomass gradually

develops even if the growth is not so fast. A fragmentation culture method available for investigating the likelihood of sporulation in *Ulva* has been presented [29]. Applying the method, sporulation can be induced in 2 to 3 days on *Ulva* thallus tissue collected from the attached population [30]. By the same method, however, the thallus blades of *U. ohnoi* and the other *Ulva* spp. sampled from the massive green tides showed no or a very low frequency of induced sporulation or took a longer time to sporulation [31]. Before *U. ohnoi* was taxonomically differentiated as a new species, this species blooming in Ohmura Bay, southwestern Japan, had been identified as a sterile mutant of *U. pertusa* (now *U. australis*) [32] and it is still believed to be so [33]. Migita [32] showed that 1 cm<sup>2</sup> thallus fragments of his *U. ohnoi* strain displayed the maximum growth rate of two-fold growth rate per 2 days in laboratory experiments at 20 ◦C, and then transplanted into an outdoor tank and grew up to larger than 1 m<sup>2</sup> in 2 months without any sporulation, while all the fragments of more than 10 wild *U. australis* thalli formed spores in the same culture conditions, resulting in the disappearance of the thallus. These results indicate that bloom-forming species have a physiological property of being less prone to sporulate, or they do not sporulate.

*Ulva ohnoi* distribute mainly in the subtropical region and are adapted to high temperatures. Therefore, it has spread to the temperate area and outbreaks in the summer. The spread of *U. ohnoi* has been increasingly reported from various regions [34–36]. Due to global warming, *U. ohnoi* may spread further into higher latitudes and cause green tides. However, in Tosa Bay, where the *U. ohnoi* green tide was first reported, this species has recently decreased sharply and instead, *U. reticulata*, which has a distribution centered in more tropical waters, has begun to increase [37]. This example suggests that each *Ulva* species has a temperature range that balances the vegetative growth and sporulation, and that individual thallus growth, or population growth, may not be possible if the temperature limit is exceeded even by a few degrees.

#### *4.2. Ulva prolifera Green Tide*

The *U. prolifera* green tides in the Yellow Sea regularly reach their biomass peak during June and July in summer (Figure 2B) [4,15]. The earliest free-floating *Ulva* patches are found in the coastal areas of the southern Yellow Sea from mid-April to early May [3,18]. These patches originally appear off the nearby rafts for purple laver (*Neopyropia yezoensis*) aquaculture, for which the coverage area is approximately 4.1 × 10<sup>4</sup> ha [3]. Annually, approximately 6500 t of the *Ulva* mass has been estimated to be released as macroalgal waste from mid-April to late-May after cleaning the *Neopyropia* aquaculture facilities [38]. In this early stage, the patches include multiple *Ulva* spp. such as *U. linza*, *U. compressa*, and *U. aragoënsis* (=*U. flexuosa* in [18,19]). However, the free-floating *Ulva* complexes move northward, associated with the seasonal monsoons and ocean currents, rapidly develop into long large bands ranging from hundreds of meters to tens of kilometers in the open sea area in late May [3], and then massive green tides are dominated by a single species, *U. prolifera* [18,19,39]. Only this species explosively grows, while the other species disappear over 20 ◦C in early summer. This suggests that the bloom-forming *U. prolifera* can continuously grow inhibiting allocation to sporulation in high temperatures, thereby differing from the other *Ulva* spp. in this region. Supporting this finding, the culture work showed that a few centimeters of the bloom-forming *U. prolifera* fragment can grow to more than 50 cm in length without sporulation at 20 ◦C (cf. Figure 12 in [40]). This growth characteristic is obviously different from that of the *U. prolifera* strain from the attached population, which cannot grow over 10 cm at ≥20 ◦C (Figure 3). However, the bloom-forming *U. prolifera* seems to allocate its biomass to sporulation around 25 ◦C, because it was observed that the green tide population began to decline at 25 ◦C from July to August [10].

Different from the bloom-forming type, the common attached type of *U. prolifera* form abundant populations on the substrate in brackish waters such as river estuaries [41,42]. Seasonal fluctuations of the largest attached *U. prolifera* population in Japan have been

described in detail (Figure 2B) [13,14]. The biomass and thallus length of the attached population regularly reach their maximum from January to March and then disappear by July. Although natural *U. prolifera* mats develop in the Chinese coast located in the south of the Yellow Sea and are harvested as edible biomass, the peak harvest is also from January to March [43], which is consistent with the fluctuation pattern of the populations in Japan. Although the attached *U. prolifera* has been used as an expensive macroalgal ingredient for Japanese dishes for a long time, its harvest has become a 'winter' tradition [13]. However, in line with recent ocean warming, the annual yield is declining [44]. It can be explained that the biomass decrease of the attached population is due to the shortening of the period when the water temperature falls below 20 ◦C. The most significant difference of the ecophysiological property between the bloom-forming type and the attached type in *U. prolifera* is whether they can vegetatively grow inhibiting allocation to sporulation around 20 ◦C or not. The seasonal difference is described in Figure 4.

**Figure 4.** Comparison of seasonal change of thallus state between the attached type, *Ulva prolifera* subsp. *prolifera*, and the bloom-forming type, *U. prolifera* subsp. *qingdaoensis*. Green and orange lines of thallus image show vegetative state and sporulating state, respectively. Green dots indicate microscopic propagules or spores released from the sporulated thalli. The two types have significantly different timings of the decline phase.

Although the issue of the massive green tide in China received a great deal of attention in 2008 [45], immediately after that in some research papers the bloom-forming type and the attached type were not distinguished, and both were confused because they formed a monophyletic clade together with the most closely related species, *U. linza*, by molecular analysis using the nuclear-encoded rDNA internal transcribed spacer (ITS) region, which is commonly used for *Ulva* species identification. However, studies of culture, hybridization, and phylogenetic analysis using a higher resolution DNA marker (5S rDNA spacer region) revealed that the bloom-forming type can cross with the attached type without any reproductive boundary, which was confirmed to be conspecific, but there are several genetic and ecophysiological differentiations (Table 1). Consistently, the other genetic analyses using inter-simple sequence repeat markers and a sequence-characterized

amplified region marker indicated that the bloom-forming type is a unique ecotype of *U. prolifera*, genetically distinct from the attached types along the Chinese coast [46].


**Table 1.** Comparison between the bloom-forming type and the attached type in *Ulva prolifera*.

<sup>1</sup> Data from the type locality population [40]; <sup>2</sup> The two types reported by [18]; <sup>3</sup> Thirty-one types reported by [42] and several more [19,43,47].

From the phylogenetic analyses of *U. linza* and *U. prolifera* using the 5S sequence, it has been suggested that *U. prolifera* had adapted to brackish water and recently evolutionarily separated from marine *U. linza* [42]. Furthermore, crossing tests suggested that the bloomforming type of *U. prolifera* completed the speciation from *U. linza* via intermediate brackish *U. prolifera* (=the attached type) because there is still a partial compatibility between the common brackish *U. prolifera* and *U. linza*, but a complete reproductive barrier exists between the bloom-forming *U. prolifera* and *U. linza* [48]. The brackish *U. prolifera* contains many regional populations that have genetically differentiated [42]. One of them may have acquired the ability to suppress sporulation and continue vegetative growth even at 20 ◦C or higher, resulting in creating the *U. prolifera* subsp. *qingdaoensis* that cause green tides. Interestingly, the *U. prolifera* subsp. *qingdaoensis* is characterized by a densely branching morphology (Table 1). This feature may facilitate the production and dispersal of large numbers of floating thallus fragments.

In contrast to *U. ohnoi*, widely reported from various regions, the occurrence of *U. prolifera* green tides has been limited to the Yellow Sea only. The special *U. prolifera* population unique to this region is fostered in the vast *Neopyropia* farm as its nursery bed, and is supplied annually as a large amount of floating mass [3,38]. The world's largest green tide appears to be caused by such a very special production cycle supported by the aquaculture activities.

#### **5. Mechanism of Sporulation**

The multicellular body of *Ulva* is composed of mostly uniform blade cells (or somatic cells) except for a small number of rhizoid cells forming a small holdfast (Figure 1). Therefore, the allocation of the individual thallus tissue to vegetative growth and sporulation is almost attributed to the blade cell fate between the alternative courses, somatic cell division, and sporic cell division. Nordby [29] hypothesized that the cell fate is controlled by changes in the concentration of sporulation inhibitors. This inhibitor hypothesis inferred that the double-layered structure of the *Ulva* thallus (Figure 1B) could be responsible

for maintaining a sufficient concentration of the sporulation inhibitor during vegetative growth. Supporting that, Stratmann et al. [49] revealed that the *Ulva* thallus produces at least two kinds of the sporulation inhibitor, one of which is a glycoprotein 'Sporulation inhibitor-1′ (SI-1), and the other is a nonprotein of very low molecular mass (SI-2). The SI-1 is present in the cell wall of *Ulva* cells and appears to be secreted extracellularly. The SI-2 is in the inner space between the two blade cell layers. The excretion of the SI-1 decreases with maturation of the thallus, whereas the overall concentration of SI-2 in the thallus stays constant throughout the life cycle. The SI-2 affects different *Ulva* species whereas the SI-1 is species-specific. Although such characteristics of the inhibitors have been shown, their molecular structures have not yet been identified.

The fragmentation culture method can synchronously induce sporulation in *Ulva* thallus tissue within 48 h, when fragmented single-layered thalli are transferred to fresh medium at a low density of fragments and cultured in optimal conditions [29]. It is explained by the inhibitor hypothesis that the inhibitors leak out from the circumference of fragmented thalli and the somatic cells that sense the decrease in the concentration of the inhibitors go to sporulation. As already mentioned, the bloom-forming *Ulva* spp. hardly, or do not, allocate the somatic cells to sporulation, even in high temperatures. Considering the inhibitor hypothesis, it is possibly thought that the bloom-forming species produce a large amount of the inhibitory substances or have an inhibitor-sensing system in which it is hardly relieved from the inhibition. The entire genome of *Ulva* has already been announced [50]. Therefore, if the inhibitors are structurally identified, it is expected that the elucidation of the switching mechanism between the somatic cell division and the sporic cell division will be achieved.

#### **6. Conclusions and Perspective**

By comparing attached populations and green tide populations, it became clear that the bloom-forming species continue to grow vegetatively with almost no spore formation even at high temperatures. However, few field surveys have been conducted on the process of the decline of green tide populations from the viewpoint of the allocation to sporulation, and future investigations are required.

Though the sporulation inhibitors were partially characterized in 1996, their molecular structures have not yet been determined. These substances are the key to determining the fate of somatic cell division or sporic cell division. It is highly possible that the causative species of the green tide have different reaction systems involving the sporulation inhibitors. Such research studies are more likely to detect minor differences when comparing taxa containing blooming strains and non-blooming strains within the same species. In that sense, *U. prolifera* would be excellent experimental material.

*Ulva* is a promising organism for carbon dioxide fixation and bioproduct production due to its high productivity [51,52]. Understanding the allocation system of vegetative growth and sporulation decay will enable greater control of biomass production and will contribute to the development of the bioeconomy.

**Funding:** This research was supported by the Kochi University research project of the Biomass Refinery of Marine Algae, the JST-OPERA Program (Grant Number JPMJOP1832), and the Wood and Cabinet Office grant in aid, the Advanced Next-Generation Greenhouse Horticulture by IoP (Internet of Plants), Japan.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Plants* Editorial Office E-mail: plants@mdpi.com www.mdpi.com/journal/plants

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18

www.mdpi.com ISBN 978-3-0365-3391-9